In this application we propose to perform testing for antibody to hepatitis C virus (HCV) on reposited serum from a large epidemiologic study of young adult crack cocaine smokers and nonsmokers, in order to answer questions about the epidemiology and non-parental modes of transmission of hepatitis C virus in this population. The results of HCV antibody testing on a small subset of reposited sera suggest that extensive transmission of HCV may have occurred among users of non-injected drugs in one inner-city community. The prevalence of anti-HCV antibody among crack smokers who had never injected drugs was quite high in one of the three communities - 26 percent, and was considerably higher than the prevalence among nonsmokers, although not as high as among the injection drug users recruited in that community (72 percent). We propose to test all available specimens from this study to determine the prevalence, incidence, and risk factors for seroreactivity to HCV. Because the study included nearly 2,000 young adults who reported never having injected drugs and collected extensive behavioral data on sexual and drug use practices, we will be able to examine a number of hypotheses about how HCV may have been transmitted in this population. We will examine at least three biologically plausible routes of HCV transmission: sexual transmission, transmission through crack cocaine smoking (e.g. through shared crack pipes), and transmission through intranasal cocaine use. Our specific aims are to: 1) Determine the prevalence of HCV seropositivity among a large sample of young adult crack smokers and non-smokers recruited from the streets in poor, inner-city neighborhoods in three U.S. cities. 2) Determine risk factors for HCV seropositivity among the participants in this study who had not injected drugs, in order to better understand nonparental routes of transmission. 3) Estimate the incidence of HCV seroconversion among non-injecting young adults in this sample.